Applied Thermal Engineering 178 (2020) 115544

Contents lists available at ScienceDirect

Applied Thermal Engineering

journal homepage: www.elsevier.com/locate/apthermeng

Bubble nucleators to enhance external pool boiling from the bottom row of a T
tube bundle
⁎
Roy S. Bartle1, Edmond J. Walsh
Oxford Thermofluids Institute, University of Oxford, Oxford OX2 0ES, United Kingdom

H I GH L IG H T S

• Passive nucleators underneath tube bundles enhance bundle heat transfer.

• Bundle heat transfer enhancement is most pronounced at lowest heat fluxes.
• Enhancement is due primarily to the early activation of nucleation sites.
• Enhancement is weakly dependent on nucleator geometry and vapour generation rate.

A R T I C LE I N FO A B S T R A C T

Keywords: The influence of vapour bubbles impacting on boiling heat transfer surfaces is examined. Passive nucleators
Heat exchangers based on boiling heat transfer were used to create gas bubbles with varying properties that impinged upon two
Tube bundles horizontal tubes above. Two nucleator devices were compared: a microscale nichrome wire nucleator, and a
Nucleate pool boiling hollow tube nucleator situated below a column of cylinders with uniform heat flux. Nucleators significantly
Bubble dynamics
enhanced the heat transfer coefficient of the lower tube in the column, which was most pronounced at low heat
Flooded evaporators
fluxes due to the early activation of potential nucleation sites by the bubbly flow carried from the nucleator by
buoyancy. Nucleator augmentation of heat transfer coefficient on the tube above was insensitive to nucleator
heat flux once nucleate boiling was established and, for the nucleator geometries tested, only weakly dependent
on nucleator geometry and vapour generation rate.

1. Introduction maximisation of cooling systems is critical [7,8]. In electronics, pho-

tonic devices require especially low temperature gradients for correct
The miniaturisation of power electronics places increased demand operation [9]. Bubble nucleators potentially offer a passive solution to
on engineers to design cooling strategies capable of dissipating high increase system power density and reduce thermal gradient.
heat fluxes. Coupled with this is the temperature sensitivity of many In a column or array of horizontal tubes, the upper tubes benefit
electronic components, demanding relatively low service temperatures from increased h compared to the lower tubes, especially at lower heat
and the minimisation of temperature gradients across the component, fluxes [10–15]. Two complementary hypotheses have been advanced to
in order to increase reliability [1]. explain this. Minns [16] suggested that a continuous liquid microlayer
Nucleate pool boiling offers high heat transfer coefficients and the is created by a vapour bubble sliding along a surface, which Cornwell
potential for passive system operation [2]. When a dielectric coolant is et al. [17,18] applied to tube bundles to explain the heat transfer en-
used, direct contact can be made between the hot surface and the fluid, hancement of upper tubes. Alongside microlayer evaporation, con-
so minimising the thermal path resistance [3]. Pool boiling on tube vective enhancement is suggested to include increased local [19] and
bundles is also relevant to shell-side heat transfer of shell-and-tube heat bulk flow velocity [13,14] and liquid mixing [14]. Similar enhancement
exchangers in low and moderate mass flow rate applications including is observed in natural convection, where heat transfer from an upper
flooded refrigerant [4], desalination evaporators [5] and kettle re- tube is enhanced by the flow from a lower tube [20,21]. Therefore, in
boilers [6]. In automotive and aerospace applications, power density pool boiling, the heat transfer coefficient is suggested [18,22–24] to be

⁎
Corresponding author.
E-mail address: [email protected] (E.J. Walsh).
1
Now at: Department of Engineering, Lews Castle College, University of the Highlands and Islands, Stornoway, HS2 0XR, United Kingdom.

https://doi.org/10.1016/j.applthermaleng.2020.115544
Received 21 January 2020; Received in revised form 13 May 2020; Accepted 31 May 2020
Available online 06 June 2020
1359-4311/ © 2020 Elsevier Ltd. All rights reserved.
R.S. Bartle and E.J. Walsh Applied Thermal Engineering 178 (2020) 115544

Nomenclature X heat transfer enhancement ratio, [-]

x wall thickness, [m]
A area, [m2]
a, accuracy [m/s2] Special characters
d diameter, [m]
f frequency, [1/s] ρ density, [kg/m3]
h heat transfer coefficient, [W/(m2K)] ψ nucleation site density, [1/m2]
hfg latent heat of vaporisation, [J/kg]
I current, [A] Subscripts
j tube number, [-]
n number of bubbles, [-] b bubble
Q heat, [W] c convection
q” heat flux, [W/m2] δ thermal boundary layer
R electrical resistance, [Ω] l liquid
T temperature, [K] s surface
u velocity, [m/s] sat saturation
w uncertainty, [%] v vapour

composed of a convective component, and an evaporated bubble auxiliary heater was situated a distance away from the tube column;
component alternating cyclically with heat transfer through a thermal preliminary experiments with 8 thermocouples distributed throughout
microlayer: the liquid pool showed a pool temperature uniformity of ± 0.15 °C was
maintained in all experimental conditions. The boiling container, into
h = hc + hb + hδ (1)
which all these components were placed, was an optically transparent
Upper tube heat transfer enhancement in a column must be due to sealed acrylic box (275 × 275 × 350 mm) that allowed the experiment
the enhancement of one or more of these components. Hence this paper to be observed by eye and photographed by camera.
examines a method to provide this enhancement for the lower tubes Miniature thermistors, electrically arranged in Wheatstone bridge
using gas bubble generation devices – nucleators. configuration, were inserted into the tubes to measure the tube surface
This paper investigates the effect of adding these bubble nucleator temperature. Type K thermocouples read the liquid pool, vapour, and
devices under a tube column. The effects of nucleator type, and power room ambient temperatures. Preliminary experiments showed that a
dissipation, on column heat transfer coefficient are investigated. High maximum room temperature variation of 2 °C across test runs had
speed imaging provides some further insight into the nucleator heat negligible effect on temperatures within the test box. All thermistors
transfer coefficient enhancement. The potential benefit of bubble nu- and thermocouples were calibrated to ± 0.1 °C accuracy in a thermal
cleation devices is considered in terms of flooded evaporator heat bath. A pressure transducer recorded gauge pressure from a tapping in
transfer performance enhancement and thermal gradient reduction the boiling container.
across temperature-sensitive electronics. The fluid used was Novec 649, a dielectric coolant with acceptable
heat transfer and environmental properties. Prior to use, the fluid was
degassed with a sonicator for 10 min to encourage cavitation, and then
2. Method and materials 45 min of vented boiling using the auxiliary heater in the boiling
container removed any remaining noncondensibles. Previous studies
Fig. 1 illustrates the experimental apparatus which consists of a with low surface tension fluids have noted that a boiling curve obtained
column of two 304 grade stainless steel tubes with heated length
by increasing power from nil can show a thermal hysteresis [25], where
93 mm, outer diameter 2.55 mm, and wall thickness 0.2 mm, were wall temperature increases until stable nucleation sites are generated,
mounted horizontally, electrically connected in series, and uniformly
and then declines as vaporisation occurs. To mitigate against this, the
Joule heated. Underneath this were placed two independently ad- boiling curves were obtained by beginning at the highest heat flux
justable Joule heated bubble nucleators for evaluation, with dimensions
tested and then incrementally decreasing the power. Three configura-
shown in the Fig. 1 inset. The tubular nucleator was simply another tions were tested: only the tube column heated, the tube column with
tube of the same specification as the column tubes. The wire nucleator
the tubular nucleator activated, and the tube column with the wire
was a 100 μm diameter nichrome wire, stretched taut and soldered at nucleator activated. Inactive nucleators remained in-situ during all
both ends into a brass solder bucket. Spacing plates to hold these
tests. To evaluate the effect of nucleator power on column heat transfer,
components were 3D printed from PTFE. A separate Joule heated

Fig. 1. Diagram of experimental apparatus with tube column detail, expanded. Dimensions in mm.

2
R.S. Bartle and E.J. Walsh Applied Thermal Engineering 178 (2020) 115544

an initial experimental programme was run with the wire nucleator and with wire and tubular nucleator q″ set at 87 and 11.6 kW/m2 respec-
a single tube above. tively.
High speed camera imagery at 100 fps, in conjunction with image Fig. 4 shows heat transfer coefficient comparisons between the no
analysis software, was used to evaluate bubble dynamics above tube nucleator, wire nucleator, and tube nucleator cases for the lower and
j = 1. Bubble diameter was measured directly from a still image at each upper tubes. As has been extensively reported, without a nucleator the
heat test condition, assuming a spherical profile. Bubble velocity was upper tube returned higher h than the lower tube for all heat fluxes
measured by measuring the vertical displacement of a bubble between tested. At q″ = 8.05 kW/m2, the heat transfer coefficient of the lowest
two images at a fixed frame rate. Ebullition frequency was measured by tube increased by 204 and 351 W/(m2K) with the addition of the wire
counting the number of frames required for a nucleating bubble to and tubular nucleators respectively, corresponding to tube temperature
depart the tube. For each test condition, 10 samples were measured, reductions of 2.07 °C and 3.09 °C.
and the mean and standard deviation reported. The heat transfer coefficient enhancement on the lower tube due to
Nucleation site density for the lower tube at the lowest heat flux was nucleator use is described as the ratio
calculated by simply counting the number of nucleating (attached)
hj = 1, with nucleator
bubbles observed on the tube and discounting any bubbles only passing Xh = ,
by, as shown in Fig. 2. It should be noted that the nucleation site density hj = 1, no nucleator (5)
and ebullition frequency measurements are subject to higher un- and graphed in Fig. 5. The trend for both nucleator types was for a large
certainty than other bubble measurements as a degree of subjectivity enhancement at low heat fluxes, caused by the early activation of po-
exists in judging whether a bubble is attached to the surface or passing tential nucleation sites, which diminishes with increasing heat flux as
over. these nucleation sites are activated by the increased surface tempera-
Heat flux generated is ture. Above q″ = 20 kW/m2 the enhancement provided by a nucleator
I 2R is diminished.
q′ ′ =
As (2)
where surface area As = πdL. The boiling heat transfer coefficient: 3.3. Bubble dynamics

q′ ′ Mean bubble diameter for the experimental conditions tested is

h=
ΔTsat (3) shown in column 1 of Table 1. For both heat fluxes, bubble diameter did
not change significantly with the use of the wire nucleator but increased
ΔTsat = Ts − Tsat (4) moderately with the use of the tubular nucleator. At q″ = 8.05 kW/m2
Kline and McClintock [26] uncertainty was computed for the ex- the use of either nucleator increased bubble non-sphericity as shown in
periment with aI = 0.3% and wL = ± 0.1 mm. Tube resistance was Fig. 6(a-c). Bubble diameter increased considerably with increased heat
measured with wR = 0.0037 Ω and returned heat flux uncertainty of flux due partly to coalescence; note in Fig. 6(c) the bubbles above the
6.29%; the wire resistance per unit length is given by the manufacturer nucleator tube are close to spherical, whereas the bubbles above tube
as 137.5 Ω/m and returns a heat flux uncertainty of 0.50%. Heat j = 1 have a profile that is more elliptical than spherical.
transfer coefficient uncertainty on the tubes varied from 6.66% to The relationship between bubble diameter and velocity in natural
7.24%. convection has been studied extensively, and most correlations pub-
lished are empirical. The Massarani correlation, appearing first in
3. Results and discussion English in Scheid et al. [27], predicts that a bubble diameter increase
from 1.03 to 1.57 mm increases bubble velocity from 161 mm/s to
3.1. Bubble nucleator heat flux 206 mm/s, which agrees well with the experimental data above for the
no nucleator cases. Ebullition frequency did not vary significantly be-
An initial test to evaluate the effect of nucleator wire power on the tween test cases.
heat transfer coefficient of a single tube above, with tube heat flux of We note that, although the addition of the tubular nucleator in-
20.6 kW/m2, was undertaken. The lowest heat flux producing uniform creases mean bubble diameter, there is no corresponding increase of
bubble nucleation along the wire length was found to be 52.6 kW/m2. vertical bubble velocity; rather the addition of a nucleator appears to
Nucleator wire heat flux was then increased up to 130 kW/m2. The heat slightly retard the bubble, especially at the lowest heat flux. As the
transfer coefficient of the tube above varied by less than 1.03% across buoyancy of a larger bubble is greater than a smaller bubble, the ad-
this range. ditional kinetic energy must have been dissipated in collisions or lateral
The effect of nucleator power was then evaluated in tube column motion; however, the horizontal along-tube component of bubble ve-
conditions graphed in Fig. 3. No notable heat transfer effect was ob- locity was negligible. Ergo the hypothesis of increased convection from
served on tube j = 2 for varying heat flux on either the tubular or wire higher bubble velocities or increased liquid–vapour mixing contributing
nucleator. Both the tubular nucleator (a) and the wire nucleator (b) to higher h on upper tubes in a column is not substantiated by these
show the same trend of a slight h increase with increasing heat nu- tests.
cleator heat flux at low column heat fluxes. As more vapour is gener- Nucleation site density is
ated, the probability that a bubble will contact a potential nucleation
site on the tube above the nucleator is increased. In numerical terms,
however, only the enhancement caused by varying the tubular nu-
cleator heat flux at the lowest column heat flux (5.09%) approaches
statistical significance, whilst the heat transfer enhancement caused by
increasing nucleator heat flux in the other conditions is lower than the
experimental uncertainty.

3.2. Heat transfer coefficient across tubes

Having established the negligible effect that varying nucleator heat Fig. 2. Image to show attached nucleating bubbles and passing bubbles gen-
flux has on column heat transfer, the results presented hereafter are erated by the wire nucleator below.

3
R.S. Bartle and E.J. Walsh Applied Thermal Engineering 178 (2020) 115544

Fig. 3. The effect of (a) tubular and (b) wire nucleator heat flux adjustment on the heat transfer coefficient of the lower tube in the column.

Fig. 4. The effects of a bubble nucleator on the heat transfer coefficient of (a) the lower tube and (b) the upper tube in a 2 tube column.

nb
ψ= .
As (6)

At q″ = 8.05 kW/m , ψ on the lower tube j = 1 increased by 31.5%

2

and 43.2% with the use of the wire and tubular nucleators respectively.
The heat removed by bubbles is
π
Qb = nb db3 fb ρv hfg .
6
Hence the evaporated bubble component of Eq. (1) is:
Qb
hb = .
As ΔTsat (8)

As the velocity data does not suggest significant increased con-

vective heat transfer, the increased h from the nucleation devices ap-
pears to come from the increased hb resulting from greater nucleation
site density. The flow of vapour bubbles from below appears to activate
and sustain potential nucleation sites. A comparison of the nucleator
Fig. 5. Heat transfer coefficient enhancement ratio showing a decrease with effect on the hb component can be made by substituting the experi-
increasing heat flux. mental data into Eq. (8). For the q″ = 8.05 kW/m2 heat flux, the wire
and tubular nucleators show hb enhancement of 307 and 990 W/(m2K)
respectively, which is around thrice the experimental enhancement
shown in Fig. 4 (a). Because bubble diameter is cubed, even a modest

Table 1
Measured bubble diameter, vertical velocity, and ebullition frequency at different test conditions.
db (mm) ub (mm/s) fb (1/s) Ψ (k bubbles/m2)

q″ = 8.05 kW/m2, no nucleator 1.03 (SD = 13.3%) 143 (SD = 7.43%) 66.0 (SD = 54.2%) 94.3 (SD = 5.11%)
q″ = 8.05 kW/m2, wire nucleator 1.01 (SD = 20.3%) 133 (SD = 9.40%) 73.3 (SD = 36.9%) 124 (SD = 8.08%)
q″ = 8.05 kW/m2, tubular nucleator 1.23 (SD = 22.2%) 124 (SD = 8.30%) 73.3 (SD = 36.9%) 135 (SD = 5.16%)
q″ = 32.2 kW/m2, no nucleator 1.57 (SD = 12.2%) 199 (SD = 7.23%) 65.8 (SD = 44/0%) –
q″ = 32.2 kW/m2, wire nucleator 1.59 (SD = 16.6%) 195 (SD = 12.6%) 65.3 (SD = 45.5%) –
q″ = 32.2 kW/m2, tubular nucleator 1.73 (SD = 11.5%) 192 (SD = 8.53%) 68.3 (SD = 38.5%) –

4
R.S. Bartle and E.J. Walsh Applied Thermal Engineering 178 (2020) 115544

Fig. 6. Images to show bubble dynamics for the three nucleator conditions at two representative heat flues.

increase in db has a large impact in hb. As hb and hδ compete for the 4. Conclusions
same heat, and as hc does not appear to change significantly, the ad-
dition of a nucleator must significantly enhance hb at the partial ex- Pool boiling of a tube column was performed with bubble nucleators
pense of hδ. However, although enhanced nucleation site density is positioned underneath. Within the experimental range, the following
accounted for in the hb term, it is the result of the generation of a was found:
natural convection bubbly flow by the nucleator device underneath the
tube. Hence convective flow effects, albeit not convective heat transfer, • Once uniform bubble generation was established on the nucleator,
can be said to account for the heat transfer enhancement produced by a the absolute value of heat generated by the nucleator had negligible
nucleator device on lower tubes in a column. effect on heat transfer coefficients in the column.
At higher heat fluxes, more homogenous nucleation around the tube • Increased heat transfer coefficients were observed on the lower tube
can be observed by comparing Fig. 6(a) and (d) as hitherto dormant as the result of nucleator use. The heat transfer enhancement was
nucleation sites are activated by the higher local surface temperatures. most pronounced at low column heat fluxes and decreased with
This has been reported before [13] and partly explains the reduced increasing heat flux.
effect of nucleator devices on h at elevated heat fluxes. • This heat transfer enhancement appears to be due to increased nu-
Producing the column h enhancement at low heat fluxes comes at cleation site activation caused by the convective flow generated by
the cost of a power input into the nucleator devices of 2.54 and 8.64 W rising bubbles from the nucleator devices.
for the wire and tubular nucleator respectively. At the lowest column • The addition of a nucleator device had negligible impact on the
heat flux tested, the tubular nucleator increases the hj=1 by 18.5% upper tube in the column.
above the wire nucleator hj=1 enhancement but consumes 3.4 times the • Nucleators are a passive solution to lower the temperature differ-
power. This limited effect of increased vapour generation rate corro- ence across the tubes at low heat fluxes, which may be advantageous
borates the findings of Section 3.1. However, because the nucleator in heat transfer processes, or the cooling of devices (e.g. photonics),
devices are operated at different heat fluxes, future work with a larger where thermal gradient minimisation is important.
set of nucleator geometries and with comparable power outputs is re-
quired to more fully evaluate the extent to which vapour generation Declaration of Competing Interest
and/or geometric effects contribute to the hj=1 enhancement variation
between the nucleator device types. The authors declare that they have no known competing financial

3.4. Temperature gradient across tubes

Fig. 7 shows the temperature gradient across the tubes

Tj = 1 − Tj = 2
∇T =
Δy (9)

where Δy is the tube pitch (or centreline to centreline distance in the

vertical y-direction). At the lower heat fluxes, nucleator use sig-
nificantly lowers the thermal gradient across the tube column; with the
tubular nucleator, the thermal gradient is reduced by one order of
magnitude at the lowest column heat flux.
The experimental setup elucidates similar physics to pin finned heat
sinks in pool boiling. In photonics devices, not only the absolute tem-
perature but also the minimisation of temperature gradients is im-
perative to maintain correct device operation because of the high rate
of change with temperature of the refractive index of silicone
(1.8 × 10−4 /K) [28]. The use of a nucleator underneath a pin finned
heat sink, cooling a photonics device by pool boiling, may merit ex-
ploration in offering the potential of relatively high heat transfer Fig. 7. Temperature gradient across the tube column for the three nucleator
coefficients and low thermal gradients across the device. conditions.

5
R.S. Bartle and E.J. Walsh Applied Thermal Engineering 178 (2020) 115544

interests or personal relationships that could have appeared to influ- [11] S.B. Memory, D.C. Sugiyama, P.J. Marto, Nucleate pool boiling of R-114 and R-114-
ence the work reported in this paper. oil mixtures from smooth and enhanced surfaces-II. Tube bundles, Int. J. Heat Mass
Transf. 38 (8) (1995) 1363–1376.
[12] Y. Fujita, S. Hidaka, Effect of tube bundles on nucleate boiling and critical heat flux,
Acknowledgements Heat Transf. Asian Res. 27 (4) (1998) 312–325.
[13] A. Gupta, Enhancement of boiling heat transfer in a 5 x 3 tube bundle, Int. J. Heat
Mass Transf. 48 (18) (2005) 3763–3772.
Funding from the Engineering and Physical Sciences Research [14] M.G. Kang, Pool boiling heat transfer on tandem tubes in vertical alignment, Int. J.
Council – United Kingdom (EPSRC Grant Number 1658609) and the Heat Mass Transf. 87 (2015) 138–144.
Future Technologies Group, Rolls Royce plc, made this paper possible. [15] R.S. Bartle, E.J. Walsh, Pool boiling of horizontal mini-tubes in unconfined and
confined columns, Int. J. Heat Mass Transf. 145 (2019) 118733.
Particular thanks is rendered to Dr. Rory D. Stieger, Rolls Royce plc, for [16] D.A. Minns, A Study of Nucleate Boiling in a Liquid Flowing Through a Vertical
industrial support of this work. Tube, PhD Thesis University of London, 1969.
[17] K. Cornwell, R.B. Schuller, A study of boiling outside a tube bundle using high speed
photography, Int. J. Heat Mass Transf. 25 (5) (1982) 683–690.
References
[18] K. Cornwell, The influence of bubbly flow on boiling from a tube in a bundle, Int. J.
Heat Mass Transf. 33 (12) (1990) 2579–2584.
[1] S.M. Sohel Murshed, C.A. Nieto de Castro, A critical review of traditional and [19] K. Cornwell, J.G. Einarsson, Influence of fluid flow on nucleate boiling from a tube,
emerging techniques and fluids for electronics cooling, Renew. Sustain. Energy Rev. Exp. Heat Transf. 3 (2) (1990) 101–116.
78 (2017) 821–833. [20] E.M. Sparrow, J.E. Niethammer, Effect of vertical separation distance and cylinder-
[2] Y. Mei, Y. Shao, S. Gong, Y. Zhu, H. Gu, Effects of surface orientation and heater to-cylinder temperature imbalance on natural convection for a pair of horizontal
material on heat transfer coefficient and critical heat flux of nucleate boiling, Int. J. cylinders, J. Heat Transfer 103 (4) (1981) 638–644.
Heat Mass Transf. 121 (2018) 632–640. [21] K. Hata, Y. Takeuchi, K. Hama, M. Shiotsu, Natural convection heat transfer from
[3] R.S. Bartle, K. Menon, E.J. Walsh, Pool boiling of resin-impregnated motor windings horizontal rod bundles in liquid sodium. Part 1: Correlations for two parallel hor-
geometry, Appl. Therm. Eng. 130 (2018) 854–864. izontal cylinders based on experimental and theoretical results, J. Nucl. Sci.
[4] S. Joo Hong, E. Seok Wang, C. Woo Park, Heat transfer characteristics of falling film Technol. 52 (2) (2015) 214–227.
and pool boiling evaporation in hybrid evaporator in vapor compression system, [22] E. Hahne, Q. Chen, R. Windisch, Pool boiling heat transfer on finned tubes - an
Appl. Therm. Eng. 153 (November) (2018, 2019) 426–432. experimental and theoretical study, Int. J. Heat Mass Transf. 34 (8) (1991)
[5] Z. Liu, E. Ishibashi, Enhanced boiling heat transfer of water/salt mixtures in the 2071–2079.
restricted space of the compact tube bundle, Heat Transf. Eng. 22 (3) (2001) 4–10. [23] J.R. Thome, Boiling Heat Transfer on External Surfaces, 2006.
[6] D.A. McNeil, K. Bamardouf, B.M. Burnside, M. Almeshaal, Investigation of flow [24] A. Swain, M.K. Das, A review on saturated boiling of liquids on tube bundles, Heat
phenomena in a kettle reboiler, Int. J. Heat Mass Transf. 53 (5–6) (2010) 836–848. Mass Transf. 50 (5) (2014) 617–637.
[7] J.L.G. Oliveira, C. Tecchio, K.V. Paiva, M.B.H. Mantelli, R. Gandolfi, L.G.S. Ribeiro, [25] M.S. El-Genk, H. Bostanci, Saturation boiling of HFE-7100 from a copper surface,
In-flight testing of loop thermosyphons for aircraft cooling, Appl. Therm. Eng. 98 simulating a microelectronic chip, Int. J. Heat Mass Transf. 46 (10) (2003)
(2016) 144–156. 1841–1854.
[8] J. Shi, D. Feng, Z. Chen, Experimental investigation on pool boiling heat transfer on [26] S.J. Kline, F.A. McClintock, Describing uncertainties in single-sample experiments,
untreated/super-hydrophilic metal foam under microgravity, Int. J. Heat Mass Mech. Eng. 75 (1953) 3–8.
Transf. 151 (2020) 119289. [27] C.M. Scheid, F.P. Puget, M.R.T. Halasz, G. Massarani, Fluid dynamics of bubbles in
[9] T. Zhang, J.L. Abellán, A. Joshi, A.K. Coskun, Thermal management of manycore liquid, Brazilian J. Chem. Eng. 16 (4) (Dec. 1999) 351–358.
systems with silicon-photonic networks, in Proceedings - Design, Automation and [28] P.A.K. Yepez, U. Scholz, J.N. Caspers, A. Zimmermann, Novel measures for thermal
Test in Europe (DATE), 2014, pp. 3–8. management of silicon photonic optical phased arrays, IEEE Photonics J. 11 (4)
[10] M.K. Jensen, J.T. Hsu, A parametric study of boiling heat transfer in a tube bundle, (2019) 1–17.
J. Heat Transfer 110 (1988) 976–981.